Vehicle Torque Allocation Using Virtual Control Commands
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The control allocation problem in hybrid and battery electric vehicles with redundant actuator suites is complex due to the excess number of actuators at vehicle corners, leading to multiple combinations producing the same virtual control, making effective torque distribution challenging.
Innovation Solution
A method using a controller to map virtual control commands to actual torque commands at each corner, prioritizing actuators based on energy efficiency and bandwidth weighting matrices, determining desired forces at the vehicle's center of gravity, and allocating these forces among corners, thereby optimizing torque control.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a redundant actuator suite is used at vehicle corners, then the system can provide multiple combinations of actuator positions to achieve the same virtual control, but the control allocation problem becomes particularly complex
Solution Approach 1:
The patent transforms the control allocation problem from a complex combinatorial optimization problem into a tractable mathematical problem by changing the parameter representation. It uses a cost function with weighting matrices that parameterize the actuator priorities and constraints, allowing the system to handle redundant actuators through continuous parameter adjustment rather than discrete combinatorial search. The cost function J(u) = (u-udes)^T Wu (u-udes) + γ (Bu-v)^T Wv (Bu-v) provides a systematic way to manage actuator redundancy by minimizing a weighted quadratic cost.
Solution Approach 2:
The patent introduces virtual control commands as an intermediary layer between the desired vehicle dynamics and the actual actuator commands. This virtual control layer (v) acts as a mediator that decouples the complex relationship between multiple actuators and vehicle dynamics, allowing the control allocation problem to be solved in two stages: first determining virtual controls, then allocating to actual actuators. This intermediary representation simplifies the overall control architecture.
2Use of energy by moving object
If multiple actuators are controlled to achieve desired forces at the center of gravity, then the system can optimize energy efficiency and bandwidth utilization, but real-time computation and coordination become more challenging
Solution Approach 1:
The patent performs preliminary action by pre-defining the cost function structure and weighting matrices that encode energy efficiency priorities and actuator bandwidth characteristics. The weighting matrices Wu and Wv are configured in advance to reflect energy consumption patterns and actuator capabilities, so that during real-time operation, the controller only needs to minimize the pre-formulated cost function without re-computing energy models or coordination strategies. This preliminary setup enables efficient real-time optimization.
Solution Approach 2:
The patent implements feedback through the cost function minimization process, where the controller continuously monitors the deviation from desired control (Bu-v) and adjusts actuator commands to minimize the cost function. The feedback mechanism uses the gradient of the cost function with respect to actuator commands to guide real-time adjustments, enabling energy-efficient coordination without requiring complex predictive computations. The term γ (Bu-v)^T Wv (Bu-v) in the cost function provides feedback on control accuracy while Wu terms provide feedback on energy efficiency.
Data Source
AI summary
A method for allocating forces among the corners of a vehicle having a redundant actuator suite includes determining a set of desired forces at the center of gravity of the vehicle, and allocating the set of desired forces among the corners of the vehicle as virtual control commands using a controller. The method also includes mapping the virtual control commands at the corners to actual or true control commands at the corners, and controlling a plurality of actuators at the corners using the actual or true control commands. The actuators may include friction brakes and wheel motors. Mapping the virtual control commands may include using a Least Squares formulation. Control of the actuators may be prioritized with respect to each other using weighting matrices. A vehicle includes a controller having actuators and a controller configured for executing the above method.